Lux et Veritas Yale Center for Perinatal, Pediatric and Environmental Epidemiology EPHLogo

Home | Faculty & Affiliates | Staff | Seminar Series | Articles | Contact Us

The Children's Air Pollution Study [the CAPS Study]

More than seven million children in the United States suffer from asthma with the highest prevalence in inner-city areas. While some cases of asthma are believed to have genetic origins, many cases are due to exposure to environmental irritants such as air pollution, pet dander, and mold.

In children, indoor exposure to air pollutants is associated with an increased risk in asthma severity including a greater frequency in respiratory symptoms and asthma medication use. Indoor pollutants can originate from a variety of sources, for example second-hand smoke and outdoor pollution from nearby traffic.

Our research project, Children’s Air Pollution Study (CAPS), will examine the role of specific residential air pollutants and their sources on asthma in inner-city children. The randomized, double-blind study design will use a modified commercial air cleaner to reduce specific pollutants in homes where the risk is highest. The air cleaner will be operated using different filters and will be in homes for a period of 18 weeks. Parents of the enrolled asthmatic child will be asked to keep a record of their child's symptoms and medication use.

We hope the results of our study will show if reductions in indoor air pollution can reduce severity of asthma. Results of our study may help suggest cost-effective interventions for reducing the burden of childhood asthma.

In the fall of 2015, we plan to begin enrolling asthmatic children, between the ages of 5 and 11 years old, living in inner city homes in Connecticut and western Massachusetts.


Family-specific Genetic Variants Contributing to Asthma Susceptibility (the FAstGen Study)

Asthma is the most common chronic illness in childhood and it is widely recognized that there is a major genetic component to asthma susceptibility. Genome-wide association studies, relying on the “common disease, common variant” model of disease, have identified common variants in more than 15 genes significantly associated with asthma or related phenotypes. While the majority of these findings have been replicated, the risk conferred by any one of these variants is low (typical OR < 1.7). Thus, there is a significant portion of the genetic component attributed to asthma development and severity that has not been accounted for, termed “missing heritability.” This missing heritability problem is not unique to asthma and has been recognized for the majority of common diseases. It has been proposed that rare variants may be an important, yet unexplored, player in common diseases and may in fact explain some of the strong associations that have not yet yielded causative variants.

Here, we hypothesize a model whereby individual families are segregating “family-specific” mutations contributing to asthma susceptibility. In contrast to the majority of rare Mendelian disorders where only one causative mutation is required for disease, we hypothesize that at least one family specific mutation is necessary, but not sufficient for disease development within individuals in the family. It is only when these private mutations occur in the context of common asthma susceptibility variants that the disease will develop.

This proposal utilizes a family-based whole-exome sequencing strategy to identify family-specific variants segregating with asthma and study the distribution of functional variants across all known asthma genes. We will take advantage of the family-based nature of several studies conducted at the Center for Perinatal, Pediatric and Environmental Epidemiology by selecting asthmatic probands from the children that had asthma phenotype information collected in one of the original studies. We propose to re-contact asthma families with three or more children, at least two of whom are asthmatic, to collect blood for DNA, update the asthma phenotype information in the parents and children, and conduct family-based whole-exome sequencing analyses. We will identify family-specific mutations using whole-exome sequencing of half of the families in the study and then perform a family-based association analysis to identify genes containing multiple family-specific variants associated with asthma. We will then sequence the genes identified in this discovery exome sequencing data in the other half of the asthma families to validate these findings. Finally, we will look across the frequency spectrum of variants in our whole-exome data to determine if asthma patients have more functional variants in known asthma genes than their non-asthmatic relatives.

Effects of Fine Particle Composition on Birth Outcomes

Preterm delivery, low birth weight and small for gestational age (SGA) are major causes of infant mortality and severe morbidity in the U.S. For African Americans, these risks are nearly doubled. Several studies have implicated air pollution, especially particulate matter <2.5m (PM2.5) as a risk factor for adverse birth outcomes. A challenge in this research is that particles vary widely in chemical composition by region and season. Due to lack of scientific evidence on which types of particles are most harmful, particles are regulated by size. We will investigate two aspects of particles’ chemical composition: 1) levels of individual PM2.5 chemical components (e.g., nickel); and 2) levels of PM2.5 from specific sources (e.g. oil combustion), to investigate the relationship between exposure to specific types of particles and birth outcomes.

We obtained filters from 8 CT and MA monitoring sites that were used to measure PM2.5 total mass. We will analyze these filters to determine levels of 49 chemical components. This chemical component data will be used in source apportionment modeling to estimate PM2.5 from specific sources (e.g. motor vehicles). Our primary outcomes will be birth weight (continuous), low birth weight <2500 gm, very low birth weight <1500 gm, preterm birth <37 wks, very preterm birth <32 wks, and small for gestational age (<10th percentile weight for gestational age), using CT and MA birth certificate data for births (2001-2006) within 30 km of the monitoring sites (N=213,000).

Associations will be assessed between birth outcomes and PM2.5 chemical components (Aim 1) and between birth outcomes and the levels of PM2.5 from particular sources (Aim 2). In Aim 3 we will examine traffic PM2.5 with three exposure methods: 1) the chemical component most associated with traffic from Aim 1; 2) traffic PM2.5 levels from Aim 2; and 3) traffic-related air pollution estimated through a GIS traffic model. In our preliminary studies, risk of low birth weight increased 8-13% per interquartile range (IQR) increase in exposure to specific PM2.5 components. We will have >90% power to detect a 10% increase in risk of any outcome (including very low birth weight, prevalence 1.5%) associated with an IQR change exposure to a specific PM2.5 component or PM2.5 source. Associations will be investigated in interaction models to test whether effects differ by race, Findings from this study will identify the components or sources of PM2.5 with the greatest impact on birth outcomes and could lead to interventions targeted at the sources most toxic to infant health.


Recent Past Projects

Traffic and Respiratory Health

Traffic, which has been steadily increasing over the past two decades, represents an important local source of air pollution, with “local” most likely encompassing a smaller area than is represented by a typical National Air Quality Network monitoring site. Environmental Protection Agency (EPA) monitoring sites are intended to represent human exposure on an urban-scale (e.g., 4 to 100 km). With the economic pressure to use more efficient fuel (i.e. diesel fuel) levels of traffic-related markers, such as nitrogen dioxide (NO2) and elemental carbon (EC) will likely rise resulting in an increased risk for respiratory effects in sensitive populations.Traffic-related pollutants, including NO2 and EC-containing particle pollution, have been associated with adverse health effects, and NO2 is a criteria pollutant monitored by the EPA. Between 1994 and 2000, over 1,100 children and 2,300 women participated in our prospective cohort studies of respiratory health conducted in Connecticut and central Massachusetts, a region including areas that have been and remain in “non-attainment” with EPA air quality standards. For analyses of the health effects of air pollution we have had to rely on the high quality, but spatially limited pollutant data measured at EPA air quality monitoring sites. In spite of this limitation, we have demonstrated adverse respiratory health effects with the temporal variability of pollution: e.g., risk of daily symptoms and/or medication use among asthmatic children with increases in daily levels of air pollution. Using Geographic Information Systems (GIS) technology and data from a variety of spatially-related sources, we propose to develop estimates of traffic-generated pollutants with higher spatial and temporal resolution than measurements from EPA monitors can provide. Our goal is to estimate levels of pollution due to traffic within buffers (e.g. a 500 m radius) around a residence within a neighborhood, i.e., residential-scale pollution levels. The primary aim of this study is to use our improved exposure estimates to examine the impact of traffic-generated pollution on respiratory health in the Northeast using our substantial archive of health information. Infants, young children and pregnant asthmatic women, groups represented in our study populations, may be especially sensitive to air pollution. The novelty of this project is the first ever use of pollutant estimates from sophisticated atmospheric models to study the effects short-term exposure to air pollution on respiratory morbidity. The strengths of this research include both improvements in exposure estimation techniques that can be applied to large epidemiological studies, and use of these exposure estimates in the analysis of respiratory health effects.

The Effect of Traffic and Air Pollution on Birth Outcomes

Preterm delivery, intrauterine growth restriction and low birth weight are major causes of infant mortality and severe morbidity in the United States. We propose to investigate the hypothesis that maternal exposure during pregnancy to ambient air pollution and traffic (a significant local source of air pollution) is associated with increased risk for low birth weight , preterm delivery and small for gestational age birth. Several studies have examined the relationships between these adverse birth outcomes and maternal exposure to ambient air pollution but results have been inconclusive. A major impediment to this research has been the large sample size needed to investigate these relationships as well as the level of personal information needed to address potential confounders and accurately estimate exposure throughout pregnancy. This study will utilize two related datasets for the same geographic area and time period: (1) an existing cohort of women followed prospectively throughout pregnancy by the CPPEE; and (2) birth certificate data. The cohort data provides well characterized variables to control for all major confounders and information about each of the mothers’ residences throughout pregnancy to accurately assess exposure. The birth certificate data provides the statistical power to investigate severe, less frequent outcomes and to examine the effects of air pollution and traffic among African American women, already at risk for preterm delivery and low birth weight. A comparison study will also be conducted to determine the reliability of birth certificate data for use in air pollution research. If an association is confirmed, reductions in specific types of air pollution may result in a reduction in adverse birth outcomes.

Indoor and Outdoor NO2 and Asthma Severity in Children


Exposure to aeroallergens and air contaminants is hypothesized to be a major factor in the exacerbation of asthma. Results of our ongoing studies suggest that nitrogen dioxide (NO2) indoors and outdoors (associated with motor vehicle emissions) at levels well below the EPA Air Quality Standard, may be associated with respiratory symptoms in children at risk for developing asthma and with the exacerbation of asthma in asthmatic children. Our data are also suggestive of an interaction between indoor aeroallergens and NO2 concentrations in enhancing respiratory symptoms in infants and asthmatic children. The potential impact is great since traffic volume is increasing and virtually all homes have significant dust allergen levels. We propose a prospective epidemiologic study of 1,533 children with active asthma (5 to 10 years of age), to test the hypothesis that carefully quantified NO2 concentrations associated with vehicle traffic on state and interstate roads are associated with asthma severity. In addition, we will determine if home levels of NO2, in the presence of common indoor dust allergens (Der p 1, Der f 1, Bla g 1, Fel d 1, Can f 1 and fungi), are associated with an increased risk of asthma severity in children sensitized to those allergens. The above hypotheses will be evaluated while adjusting for factors known or suspected to increase risk of more severe asthma, including household and school characteristics. The study population will be drawn from elementary schools in 17 Connecticut towns using a modified version of the ISAAC questionnaire to identify children with active asthma. Initial risk factors will be assessed in a home interview, and asthma severity (symptoms, frequency of ER visits, medication use) will be assessed prospectively for one year using telephone interviews administered quarterly. At the home visit, dust samples will be collected for allergen and fungi determinations and blood samples collected for antigen specific IgE determinations. Indoor and outdoor NO2 levels will be obtained quarterly for each home. Global Information Systems in combination with road vehicle density data will be used to assess the impact of traffic on asthma severity and outdoor NO2 concentrations at the home. Traffic-related pollution may represent an important environmental exposure for asthmatic children, for which effective interventions can be developed to reduce asthma morbidity.

Gene by Environment Interactions in Asthma and Allergy

Asthma and allergies are the most common chronic diseases of childhood, and their prevalence in Western countries continues to increase. Asthma and other allergic diseases are complex diseases, and environmental exposures, likely interacting with genetic risk factors, play important roles in the development of these disorders. Recent studies have shown that exposure to animals early in life may protect against asthma and allergies, and it is thought that endotoxin exposure mediates this effect. Endotoxin is comprised of soluble lipopolysaccharide (LPS) fragments of the outer membrane of gram-negative bacteria. It is postulated that endotoxin exposure may prevent the development of asthma and allergic disease depending on dose, timing, and genetics of the individual. Endotoxin is first sensed and recognized by components of the innate immune system, which then sets off a cascade of reactions, culminating in production of cytokines and co-stimulatory molecules which may be important in directing the development of the adaptive immune system. This proposal seeks to examine whether genes encoding components of the innate immune system are associated with the development of asthma and allergies, and whether endotoxin interacts with these genes in the development of these disorders. The research design will be a candidate gene case-control study, and gene-by-environment interactions will be tested. This proposal will combine resources and data from two birth cohorts: the Boston Home Allergens and Asthma Study and the Connecticut Childhood Asthma Study. In doing so, we will increase the number of cases and controls, and will be able to make use of previously collected data and specimens. Blood for genetic material is currently being collected on the two cohorts, and endotoxin will be measured from stored, frozen house dust.

Perinatal Risk Factors for Asthma in Infants of Asthmatic Mothers Study
(the "PRAM" Study)

Increasingly, women are asthmatic in pregnancy and their children are at higher risk. Yet little is known about perinatal risk factors that effect asthma development in young children. Environmental exposures (eg. aeroallergens, tobacco smoke, air contaminants) are known risk factors for asthma severity in children and atopic asthma is understood to relate to immune status, particularly total IgE response, but in complex and little understood ways. Several candidate genes are suggested which may increase risk of atopy and/or asthma. This proposal explores in-utero, perinatal and neonatal risk factors, e.g., poor maternal asthma management, genotype, immune status, lactation, and their interactions on infant asthma development, in a population of children whose experience in-utero was extensively monitored in 885 asthmatic and 1327 control mothers. We will follow 1680 infants to ascertain age of asthma diagnosis and its severity in the first 5-6 years of life. In-utero and perinatal risk factors have been obtained prospectively and include: obstetrical and delivery history, history of maternal asthma (measured by spirometry, symptoms, medication) and infections, prenatal medication history, maternal smoking history, job exposures and home characteristics (humidity, molds, pets, heating sources). The child’s immune status at birth will be assessed by stored cord blood total serum IgE. From cheek swabs, neonatal polymorphisms: B2 adrenergic receptor (Arg 16-Gly or Gln27-Glu); promoter region of the IL-4 gene, particularly at TT; Rsa-1 polymorphisms in the beta chain of the high affinity receptor for IgE (Fc PI- b ); tumor necrosis factor complex (LT a -Nco I, and TNF-308); CD14; IL-13; and IL-4 receptor will be assessed. Lactation, changes to housing characteristics, neo- and post-natal risk factors will be measured retrospectively by medical records and interview. Detailed characterization of maternal asthma and perinatal risk factors, and inclusion of black (10%) and Hispanic (19%) mothers, make this a unique "pregnancy cohort" for better understanding asthma onset in young children. Identification of pregnancy and neonatal risk factors for infant asthma may offer opportunities for early prevention.

Gene-Environment Interactions in Asthma Development

Asthma is a major chronic disease in the U.S. and the most common chronic illness of childhood. The Centers for Disease Control estimate that the prevalence of asthma has increased 75% between 1980 and 1994, with disproportionate morbidity and mortality among Hispanic and African-Americans. Genetic and environmental factors are known to be important risk factors in asthma development. The interaction of specific mutations within genes with specific environmental factors (allergens, fungi, environmental tobacco smoke, etc.), however, is poorly understood despite its crucial importance in understanding the development of asthma. In this competitive renewal application, we seek to continue to follow our population of 1,002 children (549 White, 268 Hispanic, 139 African American, and 46 Other) enrolled at birth, to test the original hypothesis that carefully quantitated environmental factors (indoor allergens and environmental tobacco smoke (ETS) exposures) are associated with the incidence of asthma in children whose health status, familial history, antigen-specific IgE status, home environment, and potential confounders are carefully characterized. We have followed 103 of the children for 3 years, 508 for 2 years and 391 for one year. During the first year of life, episodes of wheeze, persistent cough, and shortness of breath were recorded for 42, 49 and 21 percent of the children. Thus far, 180 children have been reported with asthma, with higher rates for Hispanics (Puerto Rican) and African Americans. Familial history and infant mattress dust mite levels were positively associated with risk of asthma. In continuing to follow the population, we will evaluate environmental risk factors for asthma development through age 7 and assess antigen specific IgE status of the children. In this application we are proposing to expand the prospective study to explore gene-environment interactions in the risk of asthma development. We aim to investigate the respective roles of allergens, ETS and other air contaminant exposures and, initially, one of several candidate genes – a polymorphism in the promoter region of the IL-4 gene, a gene found to be important for risk of atopy and asthma development. Blood samples will be banked so that, in the future, other candidate genes may be explored. Our extensive characterization of environmental exposures and risk factors provide a unique population within which gene-environment interactions in the development of asthma and atopy can be examined and may provide information necessary for targeting and prioritizing environmental interventions toward those with genetic susceptibility.

Paraxanthine and Reproductive Effects of Caffeine

Epidemiologic studies of antenatal caffeine consumption and adverse pregnancy outcomes have produced conflicting results. Nearly all prior studies relied on self-reported caffeine consumption as a measure of caffeine exposure. However, since caffeine metabolism and clearance greatly affect internal dose, and subsequently dose to the fetus, this measure does not provide an accurate assessment of caffeine exposure. To more fully examine the relationship between antenatal caffeine consumption and reproductive outcomes, a more precise measure of internal dose such as a urinary biomarker, is imperative. Because only 0.5%-2.0% of caffeine is excreted as such in the urine, there is considerable opportunity for error in estimating caffeine exposure using urinary caffeine levels. Paraxanthine, theophylline, and theobromine, are the primary metabolites of caffeine in humans and are readily detected in body fluids. These metabolites provide a more reliable estimate of the biologically effective internal dose of caffeine. The specific aims of this proposal are 1) to examine the association between intra-uterine growth retardation, preterm delivery, spontaneous abortion and conception delay and maternal caffeine consumption as estimated by urinary paraxanthine; 2) to examine the associations between paraxanthine, urinary caffeine, other metabolites, and self-reported caffeine intake; and 3) to develop statistical models to determine the most precise predictive factors of caffeine exposure. To address these aims, we will link the proposed detailed urinary analysis of caffeine metabolites with data previously collected on a large cohort of pregnant women (n=2478). Trained research assistants administered a baseline interview that included detailed questions on caffeine and decaffeinated beverage consumption, demographics, pregnancy history, medical history, tobacco and alcohol use, physical activity, use of nutritional supplements, and other reproductive risk factors. All women were asked to provide a urine sample at the baseline interview. In addition, women were randomly assigned to provide a sample at 20, 28 or 36 weeks gestation, along with a telephone interview. Medical records were abstracted to obtain information on obstetrical outcomes. The current proposal seeks funding to conduct detailed urinary analyses on the samples collected from this cohort. The samples will be analyzed for levels of caffeine and caffeine metabolites, including paraxanthine and theophylline.

Center Home Page | Contact Us


©2005 - 2016 Yale Center for Perinatal, Pediatric and Environmental Epidemiology  •  One Church Street, 6th Floor  •  New Haven, Connecticut 06510 USA  •  Phone: 203-764-9375  •  Fax: 203-764-9378 

Last Update: July 17, 2013